In pressure and acceleration sensors, it is desired to produce a relatively large signal power from a relatively small amount of energy absorbed from the medium. The goal is to minimize the mechanical energy necessary to produce a desired output signal. In pressure sensors, energy is absorbed from the medium as pressure deflects a diaphragm. Generally, a bar deeply notched at the center and its ends is placed across a diaphragm. Gages are placed on the plane surface opposite the notched bottoms. The strain of the bending bar is concentrated at the bottom of the notches. In acceleration sensors, energy is absorbed from the acceleration field as the seismic mass deflects relative to its reference frame. For example, a structure that is used features gages that are etched free from the substrate over an elastic hinge, a so-called “freed-gage.” With the hinge carrying the transverse load and the gages much further from the neutral axis of bending than the outer surfaces of the hinge, the gages become the most highly strained material. In both the acceleration and pressure sensor, efficiency permits high sensitivity via a small physical size.
A common approach taken by manufacturers of transducers has been to create a large field of strained surface and to place onto the more strained areas strain gages of a convenient size. Alternatively, structural means have been used to concentrate strain in piezoresistors. In piezoresistive sensors, signal is produced by changing the resistance of one or more strain-sensitive resistors excited by an electric current. Hence, in a simple plane diaphragm pressure sensor with embedded gages, much of the periphery and a broad area of the center are brought to the state of strain needed to provide signal in the gages. Although gages are placed in areas of highest strain, much of the strain energy is expended in the periphery and center areas which lack strain gages.
In a freed-gage structure only the piezoresistive material sees the full level of strain; the hinge and force-gathering structures are much less strained. Though the freed strain gage was an improvement over previous strain gages, it is still not the optimal structure to detect strain. Manufacturing tolerances impose a minimum cross section on the freed-gage; hence, for the required signal power, some minimum amount of material must be strained. The manufacturing process also imposes an upper limit on the resistivity in the freed gage, which limits the gage factor and thus, the sensitivity of the gage. In addition, heat dissipation limits the length of a device, such that the gages must be stitched back and forth across a gap over a hinge until there is enough total length to give the needed resistance. Thus, there is still a need for a stress concentrating structure that overcomes the short-comings of the freed-gage structure.